I built a startup-idea scanner. It just told me none of my 3,400 ideas are easy wins.
Spent the last 5 months building 1mil.app, a scanner that takes a market, runs live research on it, and returns 10 ranked business directions. You tell it your background, and it scores each idea on demand evidence and on whether someone like you could actually win it.
Last week I finally did the thing I'd been avoiding: I pointed the analysis at my own corpus. 339 scans. Roughly 3,400 ideas surfaced since March.
The result: zero ideas rated 7 or higher out of 10 on winnability. The best in the entire corpus was a 6.1, and it wants 1,200 build hours. 75% of everything scored below 2.
In June I took my six top-scored ideas and ran a willingness-to-pay check on them. 5 of 6 bombed. The old score measured demand and demand alone. So I rebuilt the scoring to price in competition, distribution, moat, and build cost, relative to the specific founder asking. The new model is stingy... maybe because reality is stingy too?
Here's what I learned from all that:
-
My hesitation to pick and build an idea wasn't just me procrastinating. I have been hoarding ideas for months and pursuing none, and the data confirmed why: there was no obvious winner in the pile.
-
Scores don't replace potential customer conversations. The scanner's real job is to say which idea deserves those conversations, and which nine don't.
-
A tool that flatters you is worse than useless. If my tool had done that, I'd have wasted months building something DOA.
Kinda ironic that my scanner's most useful output so far was telling its own founder "not this one, and not that one either." I shipped that behavior on purpose, and it cost me the comfortable fantasy of an easy win sitting in my backlog.
If you have a pile of ideas and can't choose, the free plan is 3 scans a month (needs an account, that's how I ended up talking to my first users):
https://1mil.app/?utm_source=indiehackers&utm_medium=show_ih&utm_campaign=launch
And if you think scoring winnability is impossible in principle, tell me why. That argument is the one I most want to have.
"3,400 ideas is wild. What made you finally pick one to build? I'm building Rallynex and struggling with that focus myself."
The insight about "distribution gap" is what really separates validation from validation theater. Most idea tools amplify the fantasy because their business model depends on keeping you building. Your tool does the opposite - it says "stop until your audience is ready."
That's harder to commercialize (nobody wants to hear it) but it's actually the most valuable signal. The 5/6 willingness-to-pay failures probably teach the model more than the successes - what matters is whether founders with distribution problems can now recognize the pattern before burning runway on a problem they can't solve with code.
Did you place successful acquisitions in it to test it? :)
The strongest product insight here is that "winnability" probably has to include one very boring variable: can this founder reach 20 qualified buyers this week without borrowing an audience?
Demand, competition and build cost matter, but for tiny founders distribution access is often the real constraint. I would score each idea against an actual founder-owned path: existing list, communities they already participate in, customers they can email, niche expertise, or paid channels they can afford to test. If the path is vague, the idea should stay low even if the market looks big.
You got that right. It's up to the user to input as much info as possible about their expertise, experience, and whatever else they'd bring to the table.
Treat winnability as a ranking model, not a truth score. The next useful test is prospective: freeze the model, take a holdout set of ideas it has never seen, and run the same low-cost validation sprint on the top, middle, and bottom bands.
Then compare precision at the top: what share of the top 10% produces a qualified conversation, a concrete pre-commitment, or a paid pilot? Also test a few low-ranked ideas to estimate false negatives. If high-ranked ideas consistently outperform low-ranked ones, the model is useful even if no score reaches 7. If everything fails equally, it is only producing persuasive numbers.
The product's moat may end up being calibration data from these validation outcomes, not the scoring formula itself.
This is the right test, and I have one early data point in its favor: re-running the same topic several times last week, the winnability ceiling swung between 1.5 and 6.0. So individual scores are noisy draws, which lands me where you did: the model should be judged on band-level precision, not any single number. Freezing it and running cheap validation sprints on top vs bottom bands is now the plan. And I think you're right about the moat. The formula is copyable. The record of which validations actually worked is not.
75% of the corpus scoring below 2 is the part that lands hardest — most idea tools tune their output to feel useful, so keeping the model stingy is what makes it a decision tool instead of a validation machine. The 5/6 willingness-to-pay failures are fascinating: did that check separate "won't pay" from "won't pay yet — no audience"? Curious whether the new scoring weights distribution enough to catch that gap.
Yeah, those are two different failure modes, and the scoring treats them separately.
Demand captures "people pay for this."
Winnability catches "but the winners had an audience you don't have."
That's the distribution gap. The classic example: ShipFast did $263k in a year, but that revenue rode on the founder's existing audience. Without it you're the
41st identical Supabase + Stripe kit nobody sees.
The 5/6 willingness-to-pay failures are the interesting part to me.
You changed the scoring model after seeing that result — what would you need to see now to know the new “winnability” score is actually better at choosing which ideas deserve customer conversations, rather than simply being more conservative?
Winnability is one more signal. The scanner already uses your background to shape what comes back, and winnability flags structural stuff like free-tier competitors or moats you can't build in code. That narrows the field. But which idea you actually pick up depends on who you can reach. A plumber and a lawyer looking at the same scan would pick different ideas because they talk to different people every day. Whether it's good signal, I'll find out from users who actually made the calls.
I appreciate you taking the time to explain your thinking.
I'd be interested in continuing the conversation by email if you're open to it. What's the best email to reach you on?
No, thanks.
Not keen on being part of your email harvesting campaign.
But thanks to your bot for stopping by.
;-)
Fair call. I can see why it came across that way.
Appreciate you taking the time to answer my question thoughtfully.